3.2 multiple regression step by step.pdf

Upload: kezia-sarah-abednego

Post on 12-Feb-2018

222 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    1/7

    1

    The Multiple Regression Step by Step

    In order to get the results of simple and multiple regressions, you need to handle with the following

    steps:

    1. Compute Mean Scores

    You need to compute the mean score of each factor in the independent constructs, as well

    as the factor of the dependent constructs, which the relationship is going to test. However,the mean score procedures of logistic regression and multiple regression are quite the

    same.

    For an example is to predict the relationship between Competence Upgrading and

    Technology Innovation. So, Technology Innovation is dependent construct, which

    consists ofOne

    factors and Competence Upgrading is independent construct, which

    consists of Two factors as named and shown in the questionnaire design.

    In order to compute the mean scores of each factor of dependent and independent

    variables, you should pick up the final formal r esul ts of factor analysis stage. This

    means that how many items still remain after the factor analysis?

    Compute mean score instruction:

    Click on TransformCompute Variable(then you will get Figure 1)

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    2/7

    2

    After you clicked on Compute Variable, you have to do some following steps:

    Look at Figure 2, you have to type factor name (you can label any name) in Target

    Variable is on the top-left panel.

    In Numeric Expressionpanel, you can type Mean (? , ? ) and question mark ?means that you have to insert items that you are going to compute after the factor

    analysis. Figure 3:

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    3/7

    3

    Or you can scroll and find the Statistical key term at the right hand side and then

    click on it. You will see Meanas shown in Figure 4. Then, you just double click on

    Meanand you can insert items in Numeric Expression panel.

    Then, you click OK. You can see the mean score of this factor on data view page as in

    Figure 5.

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    4/7

    4

    2. How to produce the results of multiple regressions?

    Analyze >>Regression >> Linear >>Select a factor of Dependent construct from the left panel to

    Dependent >> Select all the factors of independent constructs from the left panel to

    Independent (See Figure 6)

    >> Click on Statistics >> Select R squared change, Collinearity Diagnostics, Durbin-Watson, and Covariance Matrix >> Click on Continue.(See Figure 7)

    Figure 6

    Figure 7

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    5/7

    5

    For Method, you may need to select Stepwise >> Click on OK(See Figure 8)

    3. How to check and select the results?

    a. Model Summaryyou will see: R-square, Adjusted R-square, Durbin-Walson (D-W) and other

    related values.b. ANOVAyou will see: F-value and significant level (sig.)

    c. Coefficientsyou will see: Standardized Coefficient (Beta), t-value, p-value,and VIF (VIFrange, supposed that you will have two independent variables) then you may need to write down

    lowest and highest VIF value.

    d. Excluded Variables

    By following this procedure, there are some main criterions to treat the results of multiple regressions:

    1. To delete or reduce factor while running multiple regression in order to fit the rules of thumb,

    you do need to make sure and observe or compare with other factors, which have the lowest t-

    value (actually, criterion of t-value > 1.96) and p-value criteria (p

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    6/7

    6

    OUTPUTS OF MULTIPLE REGRESSION STEPS

  • 7/23/2019 3.2 Multiple Regression Step by Step.pdf

    7/7

    7

    HypothesisH9 : Competence upgrading positively influences technology innovation.

    Table 2 :The results of the influence of Competence Upgrading on Technology Innovation

    Independent VariablesCompetence Upgrading

    Dependent Variable

    Technology Innovation (TIMean)

    Model-1 Model-2

    Beta () Beta ()

    Exploration Competence (CUercF1) 0.481***

    -

    Exploitation Competence (CUetcF2) - 0.527***

    R2 0.231 0.278

    Adj-R2 0.227 0.274

    F-value 59.561 76.150

    P-value 0.000 0.000

    D-W 1.731 1.684

    VIF Range 1.000 1.000

    t-value 7.718 8.726

    Method Stepwise Stepwise

    Note:***

    p < 0.001,**

    p